The Hole of Burrowing Sponges Ljl Bioerosion·

Ocoologill (Berl.) 19, 203-21(j (1975) (C by Springer.Verlllg 1975

The Hole of Burrowing Sponges lJl Bioerosion·

Klaus Ri,ihler Department of 'nvertebrato Zoology. National Museum of Naruml History, Smithwnian IMtitution. Washington, D.C. 205G0

Heceived February 10, 1975

Summary. Among the lurge number of limestone.eroding organisms. sponge8, mninl;)' of the ramily Clionidac are of special intctellt be<:llulle of thcir efficient meHnll of subl!tmtum penetration by cellular ctching and bccUUIIC they relellsll clmrlleh::rilltiCHlly shaped calcium carbonate chipll which call be delccted in tho mud·size fraetion or lIIany 8Cdimentl'l. Tdenti· finble tTlU:'C f08lliia nud lIt..dimenhl lire of great ecological Ilnd pHleoecologicll1 significllllce. As new data on the cXCllvllliug mechanism 1111\'8 become available, the qUClltiollll of burrowing mlt:tlllnd llt..diment production 11Il\'c gained importance. E"tMlpolation from short term cxperiments (undcr 6 months) on substrate ill\'llsion are inoolll::lusive be<:aUJIll or high initilll peuctmtion mtCll fClIulling from nll.'chauical stimulation llnd lack of competition. New experimcnts show thllt the rate CUT\'C f1attclls after (j monthll lind thllt optimum long, tcrm erosion of Onco, dOCll not exCt.'(..,J 700 mg m-2 ycar-l (Cliolla lampa llnd C. uprica). Subs~nlt6 limil.fttioll.!llUld competition will furthcr reduce this rate. By monitoring the production of CnCO, chips by CIiOllU lampo.. it "'all polIIIible to link tlCtivit;r patterns to certain eUVirollnlClllt,1 factors. Medmnical stimuli, high light intcnllity. strong currents lind, possibly, low tcmperature llcem 10 11Ct.'Clerntc the burrowing procell8. Sponge.gencrated chipIJ can IIlliko up Ol'er 40% or oond mud when deposited in tho current shadow of the reef framework. Using tmnsoot counts and sponge 1lrea,·biomll/lS oolll'cl1Iion fnctol1l. the moon abundllnce of burroWing spcmgCEl on the Bermudll pl,ltform could be cn1eulnted. On suitablo hllrd bottom substmtCll it averagC6 W g dry weight per m". From thia \'lIlue the burrowing potcntilll of 2~;(j spongCEl can be catimatcd lUI g CllCOI per m' ilubstrnt:c per ycnr, Since 97-98% of the eroded limestone remnins in particulate form, the contribution of fine scdimcnlJl can nmount to 250 g m-' yearl, Attention is called to the fact that erosion mtes by burrowel1l can not dii'QCtly be oom parod with thOBO of borers or scrapers. Tho former lire intermittent and their activitiC6 lire affooted by environmentalnnd biological inlcractions, while activities of the Intter lire rather constant lind guided by the noed for food.

Inlroduction Sponges of tbe family Clionidac have long boon known for the ecological impact they cause by riddling limcstome objcct-s which thcy inhabit in most marine environments. Damagc t-o oyster cultures by clionids has been publicized since the early part of the Ninetccnth Century (references in: Clapp and Kcnk, 1963) and is still stimulating Slllyeys and experimental ftJ!lCarch (Hopkins, 1956; Hartman, 1958; Hoese and Durant, 1968).

• Thi~ work Wllij supporled h~' lhe SlIlith~onilln RCliClirch Foundation. by the Smithsonian Environmental $cicnCCll Program and by n Sydney I~. Wright Fellowship. lit the Bermuda Biological Station, Contrihution No. (j11J, .Bernllldll Biological Station. Contribution No. 10. Investigations of Milrine Shallow Water EOOllystenlS Project, Smith80nilln I.alltitution. '''' K. Riltzlcr Ecological requiremcnt.s of CliO/la, particularly salinity and temperature tolerances, competition for food nnd substrata llnd dept.h dislribul·ion were studied llnd discussed by Volz (1939), Hart.man (1957, 1958), Driscoll (1967), Bromlcy and Tcndal (1973) and Pang (1973). Possible usc of these dnta for paleoecological analyses nnd for life Ilnd post mort-em histories of shell bearing invertebrate fossils was proposed by Lawrence (1969) and Bromley (1970). The burrowing mechaniSlll used by sponges, an old controversy between advocntes of chemical Ilnd of mechanical menns has been restudied by War. burton (1058), Cobh (1069) and Rtitzlcr llnd Rieger (1973). It is now accepted that the excavation ill performed by cellular etching of the substratum, which results in freeing of characteristically shaped calcium carbonates chips Wig. I b and c). These chips are expelled through the oscula and are carried away by water currents. Although the nature of the etching agent (possibly carbonic anhydrase) is still under investigation (Hatch, 1972). it has boon calculated that only 2-3% of the eroded subslratum is removed in dissolved form (Rtitzler and Rieger, 1973). Otter (1973) and Ginsburg (1957) pointed to tho fact that sponge burrowing plays an important part in weakening of t.he coral reef framework, thus accelerating erosion by wave act·ion in shallow water. Goreau and Itartrnan (1963) demon strated much more obvious destructive effect-s caused by sponge burrowing on the fore-reef slope in Jamaica, below 30 m. They also suggested t.hat these sponges contribute to t.he large mud fraction of ee

~lateri8ls and Uethods .For the excavlltion experimenta in Bermuda pre.weighed Iceland spar blooks were attached to Cliona lampa Laubenfels (forma lampa) under water (1-3 m) using plastic coated wire. The procedures follo,,'ed those outlined by Rutzler and Rieger (1973). The blocks selected had a size clO8C to 3 X2 X 1.5 cm. In the reef crest area of ClIrric Bow Cay, Belize, slabs chiseled from fresh shells of Slrombus gifIM UnMeus were weighed and wired to Cliona. aprial Psng (forma aprka) which burrowed the coral Aeropora palmala (Lamarck) in about I-2m depth. The conch slabs ranged in size 6x4xO.5cm to 12xlOx1.8cm, After the experiments the area of initial penetration by the sponge (= contract area) and the area of total sponge coverage (ne"' gro",th) were measured using plsnimetry on pllper tracings. Then the sponge material Willi removed by soaking in commercial lIOdium hypochloride. ClIlcified fouling organisms were removed by wreful scrnping. After rinsing and removing spiculell and debris with s jet of water, the blooks were dried and weighed. To record burrowing llctivity patterns 8 sediment lraps were placed between 4- large concrete pillai'll in Ferry !'leach Entrnnoo (Bermuda). The pillai'll are fouled by a dell&6 population of the bivalve Clwmo. m~roph.ylla Gmelin, which is infested by CliQlI(J lompa. The CHOlia population pl'el>6nt above and surrounding the traps was calculated to cover 62 m!. within a 5 m radius. A fair portion of cllipll produced by these spongea could, l·herefore, be trnpped using 500 ml gla8& jnl'll with 60 mm diameter openings. To a\'oid too much contamination with resuspended sedimcnts the jarg were mounted on PVC rods, with their openings 70 em abol'e the sea floor. Two controls were illlltalled in a similar maImer, but approximately 4-0 m away from the sponge popullltion. Eaeh series of jan was removed after 48-14-9 hrg eXpollure. Those affected by sudden weather changes were discarded. In the laboratory the sediments were wll/lhed through a 200 Il-m sieve to remove large particle8, mostly plant material. All samplea were resuspended in 40 ml BCIlwater to take a representative 2 ml subllample, the rest was sllowed to settle and dry for wcighing after a short rinse in freahwater. The sub sample was membrane filtered. dried and mounted for light micl'06cope or 8C8.nning electron microscope observation. It turned out that the freshly excavated chips from dell&e Chama shell could eMily by recognized under the light micl'06cope. SEM photomicrographs were 200 K. RuttIer

Fig. la-d. Burrowing lind ,edimcnLI,roductioll by C/iOllfllwHfJ(l. (a) Spongu riddled CA/JIIIII ,,,,,urophyllll lIhcll (xO.8); (b) clchillgl:l Oil IIublllmlulll surfaCll IlIId I)(Lrtl;)' remQ\'cd (lhip (X fiOO): (e) excnnltcd lime,lolle thil' on cellulo!l(l filter (X 1000): (tI) Ae

01\1)' tUed for the confinnation or oounu Wig. Ie 8nd d). Sponge chips and other porticJe. were recorded lIel'aralcl)' b;r IlOint counting ,,,ing • Weigel graticlile and II mCQuring _Ie in the micr'OeOOJle ocular. A_mcnt. of burrowing 8ponge abundance in Bemmd/l. ,"lUI acoornpli8hcd by fmme COlillting along tmn~t lines. Elich trnlUleCt wall 2{) III long. In plllch red' IIlld olher rocky ,Irclill offllhore. their starting point and direction IOere choecn 111 nllldolll. 011 rocky OOll3tJl the dirwtion of tmllaect8 Willi perpendiculnr to the lIhore line. Uaing SCU I~A u II. m1 fnune with l()() cm" linc grid WlllI l>ollitioned cvery 2 m on tho tnmaect lind the bottom undcrnClHh wall carefully lICarched fol' Crlllllll or pupilille of c1ionitillpongell. The cxtent. of the Iltlpillllry field8 WIUI eIItinulI.t.,<1 to the e1Ol1Clit II. of Cflch grid frllme (5 X;; cm) /tnd recorded for cltch 8pccics (all defined by Biit1:ler. 1974). Of the onl~' lIon·cliouid burrower. S,iheckMpongi/,l othdl/,l Ilone but. the ~'oung exca\'llting 8tagCll were notcd. Biomll8ll eillimatcs for the diHcrent apecieIJ and foml8 are hued 011 " 8uhanmp1cll each. After D1ea8uring the aurflloo area. of the papillll(l fickb or of the ec::loeomal c:rtl8U. the 8pongee "'ere treated with changell of dilute h,}'drochloric:: add until tbey were completel:r dealcified. Care "-118 taken not to 10lIe any ti.-ue fragmelll8 b}' enclOlling the $l.lUples in bolting cloth. ~ndolithie algae lind other organialll8 ....-ere relllo\'OO arIel' decalcification while the ti!lluee were auspended in .....nler under II dissecting miel'OeCOpe. Wet weight. was taken after quick blotting. dry weight. after drying 10 CQn8tunt. weight of IOSo C. Determination of llIIh.free dry-weight...... IUI not considen.·d 8ignificnnt becuulIO of the fnirl}' conllistent. spicule to 110ft til!8ue mtio in nllllpecic8. The Role of Burrowing Spongee in Bioel'Oekm 207

lteslllls Long Tum Excaootimt Ralu and Subatrah Availability The Iceland spar blocks used in the Bermuda experiment6 have the advantage of being dense, pure and homogenous calcite. The well defined crystal surfaces provide maximum contact area for attachment to sponges and can be easily cleaned from unwanted fouling organisms (t!.g. crusto8e coralline algae). Neu mann's (1966) observation that the substrate mineralogy has little effect on burrowing rate could be confirmed by comparing rates in Iceland spar with those in Chama shclls. Porosity of the S\l bBtrnte aids quicker and decper penetration but no more material is removed per unit time. Porous calcarcnite also poses experimental problems because of mechanical crumbling and invasion by other organisms. Since calcite blocks could not be obtained in large sizes, s.1abB of fresh conch sheUs were used in Belize. These are an important naturally occurring sub8trate for c1ionid sponges. Both species of sponge studied have an incrusting habit which facilitates quantitative assessment and preventa mOElt 8OOl\ndary invaderil from settling. Only some algal felta, cruslo6e Corallinacea, Foraminifera (Homotrema), hydroids and tunicates (ClaV4!llina) settled on parts of the substrate blocks, and $Orne blue.green algae penctrated superficial laycu. The competitive advantage of the incrusting habit is demonstrated by the largc surface areas (in the order of square meters) these spongea can occupy. Large coral lu:ads have probably been killed by spreading CUOM wmpa bceause the delicate surface structurc of the corals (e.g. Sidera8trea, Diploria, Monla8trea) is still distinct undcr thc sponge crust (Rlitzlcr and Rieger, 1974). When the corals had been dead and exposed for only a few weeks filamentous algae had taken over and obscured the structure. This is supported by labeled and photographed Jive specimens of Monla8tna annulari6 bordering CliQna tampa in Harrington Sound, Bermuda. During a 9 month observation period (August 1973-May 1974), the sponges advanced 0.9-2.1 cm at the expense of the coral (R11tzler, unpublished). Glynn (1973) arrived at similar results (2 em per 9-month period) by transplanting Clion4 sections to injured but uninfested sites of Sickrastrw. Large coral specimens, mainly Acropora paimala can be entirely covered by CliOM aprlca near the Carrie Bow Cay reef crest. Table 1 list6 the resulta of field experiments ranging from 33 W 350 days, including the data of Neumann (1966). The area of cont-act (attachment) with the "mother" sponge is the only measure of biomass available for the cxpcriment-s up to 4-5 months. If this contact is broken, the wound heals forming ectosomal tissue with ostia. and oscula undistinguishable from naturally grown specimens. On substrata older than 5 months ectoeomal tissue bll8 already 8pread from the sides and from point6 where the sponge bll8 penetrated through the upper substrate surface. This growth can be considered a new sponge specimen with it6 own independent aquiferou8 system. To ita measured surface area the bUTfO\\'ing rates can be related. Rates related to the initial contact area only, reflect. the growth potential of the mother sponge but have no bearing on the sponge biomll8ll involved (Fig.2a). The means follow a power regression model (y=63.8zl •4) with a coefficient value r" = 0.93. Relatoo to actual surface coverage values the 208 K. Riitzler

Table 1. Field experiments on tho burrowing rate of Cl~II4Ia".pa in lcel.nd lpar (Bermuda) and of O. apr1ca in conch abell. (Carrie Bow Cay). z= mean ± ltandard error

Experi. Dura· ~)rif' Total Co,· Sponge Weight Weight ment tion ,," w!light "~ 1088: loee: No.- 1, weight ,~ "'""~ (em') mg cm-' rog em-I day. ,,1 ,,1 (eml ) contact sponge N 4/S012 33 6.30 0.77 7.66 7.66 .. .. $011 7.80 0.27 7.301 7.34 30 30 LBI3 12.59 0.28 10.34 10.34 27 27 i_ M±23 1_ M±23 N 6fBLl7 '00 17.00 3.66 ...., '.66 66' 66' m,l22 ,... 2.18 4.21 ".21 SG7 6.81 2.32 '.M '.M '"

B 220jA 220 17.92 3.22 8.' 663 383 B 13.31 6.52 •••5.3 10.9 1230 ,.. C 10.64 '.45 ,.. 3.6 66' 66' D '.50 3.6. .~ ,.• 813 63' <- 807± 155 :i-=573± 66

B 312/A 31' 13.71 6.39 '.3 7.' 1488 900 B 7.96 • .07 3.8 ••• 1071 .11 C 10.97 '.00 ••• 12.2 ... 36. i_II58±170 i-560± 170

CBC350/A 350 296.7 111.4 36.6 104.6 3'" .006 B 121.1 37.6 20.4 109.6 '838 34' C 226.6 89.9 26.9 127.4 3342 706 D 102.8 27.2 51.0 .94. .00. E 80.' "28.9.• 23.0 49.6 1257 ,. 70.4 16.6 20.8 42.7 793 '"386 G .... 19.5 37.4 2167 621 H ".3 18.5 30.'••• 25.0 .., 740 f_1877±348 f_673±96

• N _Neumann (1006), Table I. B _Bermuda. CBC_Carrie Bow Cay, Belize. The Role of Bunowing Spongee in Bioeroeion

2000 a

1500 j 1000 I I '" • • , , • • • , • ." II 12 molllh, mil em-, '"

." b

00' I '" I '" '" '" '", 2 3 .. 5 • T I II 10 11 12 IIIoftlh. 2b Fig.2a Ind b. Bunowing ratel of Clw- (mg CaCO, per em' lJP01l8I!l) rellted (I) to initill lPOnge4Ubetrate oont&ct lrea, (b) to finallponge .urfaoe area. Data point! repreeent meal\l± I.e. Open Iymboll: C. lampa; blaek Iymbol.: C. 1Jpr1ca (furt,her explanation in text) rate curve is steep aft-er an initial adaptation period of 1 month but flattens after about 6 months (1<~ig. 2b). The illustrated regre88ion curve results from the logarithmic model y = 131.7 +236.3 In :t (1'1= 0.80). It can be assumed that after approximately 6 months the stimulation of the new substrate has ceased to be effective and that substrate limitation and competition for space and food retard the activity. Some energy will also be diverted from the burrowing process to nutrient storage and reproduction. The low mean value of the 220 and 312 days experiments illustrate the im. portance of substrate availability. The blocks used were too smaU for the long exposure time and were alm08t entirely overgrown and excavated to a high

U Oeorlloe\& (Bert.), Vol. II 210 K.. ROuler

Table 2. Weight. I~ of ClioM 16",1'0 infested C1ltJfNJ. shelle after 220 days. Experiment. C OOflllieta of 3 cemented shells

Experi. Shell + Penetra- Caco, Sponge CaCo, Caco, ment sponge wet. tion weight (g) area weight (g), 1CNJ8: weight. (g) e6timllle before (from (ern Z) nfler mg em-I bcfol'Q (%) 8ubllampl1l6) 220dll.y8 sponge

A 54.6 23 50.1 24.5 39.3 441 B ..,.. 18 40.1 22.4 38.' .. %=263

C 152.8 38 134.9 71.1 130.9 56 D 62.7 44 54.8 27.3 50.8 154 i_ll)lj degree. When no morc spreading is possible burrowing continues to a certain limited depth and until about 50% of the Bubstrate is removed. This value has never boon exceeded in the two species studiod, neither in experimcnt.s nor in collected specimens. Other clionicls are known, however (!!i.g. Clion« ulala), which eliminate their substrate entirely and become a massive sponge in 11 later gro,,1.h (gamm~l stage. Maximum penetration depth measured was 2 em in compact substrata. This limit could be due to difficuJties in removing excavated chips from greater depth. Very porous calcarenite and corals can be infiltrated by Cliona lampa to a depth of 8 em. In solid calcite, therefore, I em! sponge can erode a maximum of I em! (50% removal), which equals about. 2.7 g, un.less scraping predat-ors arc aiding the process. An additional attempt. was made to obt·ain figures on long.term rates by measuring the weight change of Cliona specimens which were already well estab lished in shells of Chama flIaarophylla, after 220 da}'s. 1'he upper surfaces of the shells used were entirely covered by the sponge but the substrata showed different degrees of penetration. The main BOurce of error in tills experiment is that the weight of the shell without sponge at the beginning of the experiment has to be calcuJated from 8ubsamples which were macerated from sponge tissue. The result-s (Table 2) show a mean loss of 263 mg CaCO! per cm~ for the wenkJy penetrated shells, 105 mg Caco~ per cm~ for the heavily penetrated shells, per 220 days. This corresponds to I\. burrowing rate of 174-436 mg em-1 yearl . lh'Om these daLn. and from l~ig. 2 b, it seems safe to assume t.hat. t.he long.term mean burrowing rate of the clionids studied docs not exeeod 700 mg em-I ycarl, even considering the greater activity of freshly recruitod populat.iolls. Of this amount 97-98% (Rtitzlcr and Rieger, 1973) are released lUI fine sediments, in addition to larger particles resulting from secondary breaking down of the weak· cned substrate by wave action and scraping fishes and invertebrates.

Activily·Facror Rtlalionshjp8 It has boon shown above that substrate availability and competition are the main factol"ll controlling long-term excavation rates of Cliona. There are, however, other ellvironmental stimuli which can at lenst. temporarily affect the burrowing activit.y of these sponges. Tho Role of Burrowing Sponges in Bioerosion 211

Table 3. Sire-frequency count.., of sediments trapped at Ferry Reach Entranoo during different temperature and wind conditions Sample Month Experi- Exp06uro, Mean Mean frequency (%) of fine sediment.., acriClJ 1-· m~t wind sedi- wmper- duration direction menta. Total Cliena ehips only aturo) Ihl tion §. [ "to §. [ §. [ t ~ ~ ~ ~ (mg cm- 11 ~ ~, day-I) ... 1 ~ 1 a ~ ~ a ~ 4JI-8 January 48 +-N 7.7 18.1 26.8 55.1 2.5 11.8 2.3 (7l.=8) (16-17° C) (light-moderate breeze) Control 6.' 15.7 21.4 62.9 L1 1.5 6 12J1.-8 48 ++NW 24.3 16.7 30.2 54.1 2.2 13.5 6.1 (7l.=6) (stiff breeze- moderate gale) Control 36.2 26.6 22.4 51.1 1.4 ,., '.1 IOAI-7 August .2 1.3 10.2 28.8 61.0 0.2 3.2 1.8 (n=7) (28-29° C) (calm) Control 2.8 6.7 22.4 70.9 6 6 0 16At-8 144 +N 2.7 14.3 37.7 48.0 ... ,.• '.7 (n=8) (stiff breere) Control 1.7 18.3 30.3 51.6 6 6.7 0.'

It was casually observed in the laboratory t.hat cutting some sponge tissue from experimental substrate blocks resulted in stimulation of the burrowing process. A similar reaction is probably triggerod in nature by scraping organisms. Likewise, illumination of aquarium specimens by a focused low voltage micro scope lamp caused previously inactive sponges (6 cm2 surface area) from 7 calcite blocks to produce 6 mg of chips within 48 hours. This corresponds to 0.5 mg cm-2 day-I, the highest rate obtained under laboratory conditions. It would take more work to evaluate light as an ecological fuctor but it is known that Oliona lampa (forma lampa) only occurs in fully illuminated shallow water environments (Laubenfels, 1950; Rtitzler, 1974). Strong currents in the aquarium also seemed to have a stimulating effect which, however, was difficult to qUl\lltily due to the unavoidable dispersal of chips. In situ sediment traps near large Oliona populations, as described in the introduction and methods sections were, therefore, employod to estimate the burrowing activity relative to water movement. Since sample acries arc available from both the coldest (January) and warmest (August) season, it is possible also to deduct information 011 the possible influence of tempcrature or of pheno mena. related to it. A summary of the sampling results is given in Table 3. Particle counts are restricted to the 125-16j.Lm (3-6 phi) size rangc in which sponge.derived frag ments are clearly recognizable (see methods section). Fig. 3a shows a complete size-frequency histogram for pure sponge chips obtained from OU01UJ, larnpa ...infested Iceland spar in the laboratory. 212 K.Riitder " " -• " • • • 3 4 5 .3 .. 5 •

Fig. 3a and b. Size-frequency hiatograma of (a) pure'" Iceland apar chipll produced by Cli

Jan u a r y

"• " ". .. ". •• • •, • •

August

"• - " •• ,. ,,' • •, • • 0001'01 , • •• , • • • 3 .. 5 • 3 .. 5 • phI It E- It Et Fig. t. SiZ&-frequeocy histograms of trapped eedimenu reflecting crimm Ultivity·factoOr relationahipll. Proportion of .ponge-generatedchipll(hatcbed) isgiven in percent., temperature, E expoaure to water movement.

The histogra.ms (Fig. 4) illustrate the results from Table 3. It is obvious that strong water movement stimulates the burrowing activity (winter and summer). This is consistent with the preference of the species for habitats with exposure to fast currents (Laubenfels, (950). There are al80 indications that low The Role of Burrowing Sponp in BWefOllion 213 temperature might favor the activity but data at hand are not oonclusive. Firat, the"calm" winter period sampled was still fairly agitated and, even more im portant, had been c10eely preooded by a period of storms. Furthermore, preliminary observations on the reproductive cycle of Cliona tampa (RUtzler, unpublished) indicate that the peak of reproduction coincides with the cooling of seawater in autumn. In this case invasion of new substrata. could account for an increase of the burrowing activity. The highest percentage of sponge chips obtained by fJOdiment traps (22%) i.s much lower than that reported from sediments in the Fa.nning Island lagoon where sponge-derived particlell make up 30% of all fractions (Fiitterer, 1974). There, however, we are dealing with a low energy environment where 80rting by currents accounts for the ooncentr&tion of very fine materials. On a smaller scale, similar conditions can be found even in high energy reef environmenta, like the Carrie Bow Cay roof crest. There, the coral framework providea for almost stagnant water bodies where mud pooketa can develop which remain undisturbed, particularly as long a8 the main wind direction remains collfltant. A sample from one of these accumulations of reef mud contained 41 % sponge chips in the 125-16 fLm size range (Fig.3h).

Sponge AbuM411U and E%£4.mticm RaJu To elltimate the impact of hioer'06ion and eediment production on 8. particular marine environment, it becomes neeeasary not only to know the activity rates but also the abundance of the organisms under consideration. With the focus on clionid sponges, this point has hitherto been neglected. There is no doubt that the procedures involved are tedious and that they lack great &Ocuracy. First of all, most species of clionids are difficult to detect because of their cryptic habit, with inconspicuous papillar groups usually hidden under algal turfs. Secondly, collecting of whole specimens involves considerable physical effort including chiseling under waur and lifting heavy loads of rock. For & distributional sponge survey, Sari (1966) hll.8 applied a. frame ooanting method by which al80 the extent of clionid papillar fields can be quickly evaluated. for quantitative comparisons. This method h&8 been adopted. here with the following modification. Since several species with different burrowing patterns and substrate penetration capabilities are to be oompared. a factor had to be found to oonvert the surface area of papillar fields or incrustations (as in Olimul lampa) to biom&8ll units. On a comparative basis these units are beat Buited to reflect the burrowing potential of a sponge. The conversion faotors for tho 8 species studied were obtained from dooalcified. subsamples and are listed in Table 4. The main variance of the samplell is due to different penetration depth. Repetitive samples of t·he same specimen had standard errors of less than 10% of the mean. Table 5 showB the results of 18 hard bottom transect oounbl from II localities on the Bermuda platfonD. The average value of 16 g (dry weight) burrowing sponges per m' is oonsidered to be representative for near shore patch reefs and rock bottoms which do not remain burried by sedi.ment.a. More data are still needed. for the deeper off-shore reefs. Exceedingly high oonccntrations of Cliona tampa can be observed at Ferry Reach Entrl\nce (North Shore) and at the Flatta channel leading into Harrington Sound. There the biomass of burrowing sponges increa8C8 by 12-14 times (185-235 g m-I ) for areas of 10(1 ml , or more. 214 K. Riltzler

Table4. Relation between visible sponge area (fields of papillae or incrustations) and biom!l88 lor Bermuda burrowing spongea

Specice Mean wet weight±8.e. Mean dry weight ±H.e. (mg em-I) (mg em-I)

CIWna ooribboea 144.7± 38.8 18.9± 5.8 CliQna Ilavilodi714 45.5± 28.5 7.6± 4.7 Cliena paw;i8pintJ 838.8±188.8 t37.5±27.6 Oliona vermifera 138.4± 23.5 25.6± 3.8 Oliona dwry88a 146.1 ± 34.3 22.2± 3.9 Cliona lampa I.lampa 30B.3±153.7 43.3±15.7 Cliona lampa f.occulta 42.5± 5.4 5.2± 0.9 Cliona amplil;at'ala 240.7 ± 86.6 30.1 ±12.2 Sphecio8pongia ot~lla 215.0±198.2 28.8±21.7

Table 5. Reaults of tranaect oountB, indicating density of burrowing sponges in

TranllCCt Location Substrate Depth (m) Cliona O./l«t'i- No. carib- lodina """ Ta Castle Barbour Co~l 0.5-1 9.3 0.08 b Coral 2-3 11.0 0.3 , Calcarenite 1 2.1 II Bailey8 Bay Flaw Coral 2-3 4.9 0.09 ill Baileys Bay Caloonite'llhell 1-2 2.5 0.07 IY, Three Hill ShOlLls • Coral 2 0.' 0.5 b Coral 5-

It is only an assumption, although reinforced by the Carrie Bow Cayexperi mont, thatthe burrowing rate potential is of similar magnitude in different species. If this is the ca80 the calculated maximum long-term exC&vating rate of 700 mg 2 per cm sponge per year can be converted to 16 mg Caco3 per mg sponge (dry weight) per year. Hence, for average Bermuda rock substrata suitable for Cliona colonization these sponges could be directly responsible for the erosion of up to 256 g m-I yearl , which corresponds to a rock layer 0.1 mm thick. The output of fine grained calcarenite from this activ.ty would amount to approximately 250 g m-2 yearl • In areas of particularly high sponge concentration the erosion figure can increase to about 3 kg m-I yearl (1 mm per year), thus accounting for formation of local topographical features, such as subtidal notches, as 8Ug gested by Neumann (1966). The Role of Burrowing Spongee in Bioerosion 215

Rate comparisons "ith other rock.destroying organisms must be made with care because of theapparent differences between burro\\ing and boring orscraping organisms. The fonner produce excavations to live in (Carriker and Smith, 1969). Their activity fluctuates greatly and depends on substrate avaiJability, environ· mental conditions which influence growth and reproduction, and on biological interactiOIlB (competition, predation). Bacteria, fungi, algae, lichens and most invertebrate animals belong to this group. On the other hand, borers like many mollusks (Carriker and Smith, 19(9) and scrapers, like aome mollusks and oohinoids (Krumbein and Van der Pers, 1974) and Illany reef fishes (Bardnch, 1961; Gygi, personnl communicntion) are mobile and constnntly netive because their food supply depends on boring or rasping of calcareous objects. Such dif fcrences should bo kcpt in mind when overall rates of bioerosion and sediment production on limestone coasts or in coral reefs are being estimated. auitablc .ublltrata on the Bermuda P!aUorm. (g dry weight per ml .ubllt...te area)

C. QU7lIIi· O.dio O.tampa C.tarPIpa O.a",pli. Tou.l /~ ..... f. tampc1 f. oet:1ilIa w

2.3 ,.\ 2.1 2.0 28.3 Ir;.t 1.8 1.6 29.8 1.0 8.0 '.1 0.' OA 19.9 I.' 0.1 6.6 0.' 1.6 8A 0.09 6.5 14.8 3.\ 6.2 '.3 0.' 0.3 3.3 3.' 3.\ 6.' 0.01 '.3 t.8 6.' 6.' '.6 6.' t2.1 2.0 3.' 30.8 t.8 0.' 0.' 24.6 ,.\ ,.\ 1.1 15.7 '.2 0.' 8A 11.0 0.6 0.' 0.' 12.2 0.\ 0.\ 0.2 17.7 '.0 \.8 ,.• 66.' 0.' 6.2 OA 6.' 0.3 ,., •. \ 2.2 2.6 0.6 '.0 10.1 i=16.0

Ackl\Owltdgl1l~Il"'. I gralefully acknowloog6 dil8CUl18ion with Dr. O. Bret./lChko, and Wat.. ney's ned Barrel for many great. thoughte.

References Budach, J. E.: Tranlport. of calcarooWl fragment, by red fishe.. Sctenoe 133,98--99 (1961) Bromley, R. G.: Boringl" moe f08lli.la .nd Eldobia crtloua Portlock," an example. In: T. P. Crimes, J. C. Harper, ed!I., T...oe fc:.ik. 0001. J., Speeial J.ue 3, 49--90 (1970) Bromley, R. G., Tenda!, O. S.: &le of .ubet...te competition and phototropism bet"'een two c1ionid llpongetl. J. Zool. Lond. 16t, 161-155 (1973) Carriker, M. R., Smitb, E. H.: Comparative calcibiocnvitology: summary and conclu8ion" Amer. Zoologil:lt 9, 1011-1020 (1969) 216 K. Riitzler

Clapp, W. F., Kenk, R.: Marine borel'll, an annotated bibliography, 1136 pp. Washington, D.C.: Offioo of Naval Research, Department of the Navy 1963 Cobb, W. R.: Penetration of calcium carbonate 8ubetrates by the boring sponge. Cliona. Amer. Zoologist. 9,783-790 (1969) Dri800lI, E. G.: Att&ebod epifauna'8ubslrstfl relations. Limnol. Ooeanogr. 12,633-641 (1967) Fiitterer, D. K.: Significance of the boring sponge Gliena for the origin of fine grained material of carbonate eedimente. J. sediment. Petrol. 44, 79-84 (1974) Ginsburg, R. N.: Early diageneai.B and lithification of &hallow-water carbonate eedimentll in south Florida. Spec. Pubt. Soc. Eoon. Paleonoologiet.s and ~tineralogi8taIi, 80-100 (1957) Glynn, P. W.: Aepecte of the eoology of coral roofe in the western Atlantic region. In: O. A. Jones, R. Endean, edll., Biology and geology of coral reefa, vol. 1, Biology, p.271-324. New York·London: Academio Prees 1973 Goreau, T. F., Hartman, W. D.: Boring eponges aa controlling facton in the formation and maintenance of coral reefe. In: R. F. Sognn&ell, ed., Mechanieme of hard tillllue de etruction. Publ. Amer. AlIlln. Advmt. Sci. 75, 25-54 (1963) Hartman, W. D.: Ecological niche differenciation in the boring epongea (Clionida.e). Evolution 11,294--297 (1957) Hartman, W. D.: Natural hietory of the marine epongee of 80uthem New England. Bull. Peabody !lIue. nat. Hiet. I!, 1-155 (1958) Hatch, W.: Phyeiology of ehell penetration by Clitnul. Wal{l (Abetract). In:!l. R. Carriker, ed., Ninth and Tenth Annual Reports on Progtel:l6 and Ten Year Summary, Syet.ematice. Ecology Program, Woode Hole, MlLlIlllIrChueetUl, p. 13-14 (1972) Hoeee, H.D., Durant, J. E.: Not.ea on the boring epongee of Georgia. Completion Report. Project 2-IO-R. Ga. Game and Fieh Comm., 7 pp. (1968) HopkiDII, S. H.: The boring epongee which attack South Carolina oYlltera, with notee on lJOme lL880Ciated organillme. Contr. Bean Bluff Labe. 28, 1-30 (1956) Krumbein, W. E., Van der Pel'll, J. N. C.: Diving inveEltigatioiUI on biodeterioration by lMJ&,·urchil\lj in the rocky eublittoral of Helgoland. Helgoliinder wi88. Mooresuntera. 26, 1-17 (1974) Laubenfele, M. W. de: An ecological diBeullBion of the epongee of Bermuda. Trans. Zoo!. Soc. Lond. 27, 155-201 (1950) Lawrence, D. R.: The Ul!El of olionid epongea in paleoenvironmental analy_. J. Paleont. 48,639-543 (1969) Neumann, A. C.: Prooe88e& of recent carbonate lIed.imentation in Harrington Sound, Bermuda. Bull. Marine Sci. 15,987-1035 (1965) Neumann, A. C.: Obeervatione on ooaatal eTOllion in Bermuda and meaauremenUl of the boring rate of the eponge, Clitnul. lampa. Limnol. Oceanogr. 11, 92-108 (1966) Otter, G. W.: Rock-destroying organieDl8 in relation to coral reefe. Soi. Rept. Great Barrier Reel Exped. 1, 323-352 (1937) Pang, R. K.: The ecology of eome Jamaican excavating eponges. Bull. Merine Sci. 28, 227-243 (1973) Rutzler, K.: The burrowing epongea of .Bermuda. Smithllonill.n Contr. Zool. IGli, 1-32 (1974) Rut·zler, K., Rieger, G.: Spon~ burrowing: Fine etruoture of Clitnul. lampa penetrating calcareoue eubetrata. Marine Biology 21, 144-162 (1973) Sara, :l'ar.: Studio quantitativo della dilltribuzione dei Poriferi in ambienti euperficiali della Riviera Ligure di Levante. Arch. Oceanogr. Limnol. 14, 365--386 (1966) Volz, P.: Die Bohrschwiimme (Clioniden) der Adria. ThalaBllia 8, 1-64 (1939) Warburton, F. E.: The manner in which the eponge Clitnul. bores in calcareoue objec~. Canad. J. Zool. 88, 555-662 (1958)

Dr. Klo.U11 Rutzler Department. of Invertebrate Zoology National MUl!Elum of Natural Hietory Smithllonian Inet.itution WBilhington, D.C. 20560, USA